New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 678687 link

Starred by 3 users

Issue metadata

Status: Fixed
Owner:
Closed: Feb 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 1
Type: Bug


Sign in to add a comment

Mash browser_tests need to run more tests

Project Member Reported by sky@chromium.org, Jan 5 2017

Issue description

Currently mash browser_tests run a single test, they need to run as many browser tests as is possible.
 
Cc: jonr...@chromium.org sadrul@chromium.org

Comment 2 by sky@chromium.org, Jan 9 2017

Cc: -jonr...@chromium.org
Owner: jonr...@chromium.org
Status: Assigned (was: Untriaged)
Some more details:

. At the time I enabled browser_tests to work with mash we wanted ash and the ui to run once, and all tests run in that environment (think of ash as the system shell). That isn't realistic with the way the code is structured though (ash expects singleton connections from chrome, and each test would try to provide these). There is likely a bunch of simplification that can happen now because we start ash/ui per test.

. We didn't want to build a separate executable. This was important when targeting multiple platforms, but now that mash is chromeos specific that isn't important. If it simplifies things we should create a new test target. That said, a single target is preferable.

The code for this is here: https://chromium.googlesource.com/chromium/src/+/master/chrome/test/base/mash_browser_tests_main.cc and https://chromium.googlesource.com/chromium/src/+/master/chrome/test/base/mojo_test_connector.cc .
It would be nice to have some docs about how to run mash browser_tests locally. There's no mash_browser_tests executable (by design) but running with the same flags as the bots doesn't seem to work locally.

Bots gn flags:
dcheck_always_on = true                                                         
ffmpeg_branding = "ChromeOS"                                                    
is_component_build = false                                                      
is_debug = false                                                                
proprietary_codecs = true                                                       
symbol_level = 1                                                                
target_os = "chromeos"                                                          
use_goma = true                                                                 
use_ozone = true  

Optional alias to put in front of test calls

alias hide_tests='xvfb-run -s "-screen 0 1024x768x24"' 

The trybot runs the tests via:
./out/${OUT_DIR}/browser_tests --test-launcher-bot-mode --override-use-gl-with-osmesa-for-tests --run-in-mash --service-overrides=chrome/app/mash/mash_service_overrides.json --test-launcher-filter-file=testing/buildbot/filters/mash.browser_tests.filter --use-test-config --test-launcher-summary-output=/tmp/something_something_something.json

You can run as not bot mode for more gtest controls:
/out/${OUT_DIR}/browser_tests --run-in-mash --service-overrides=chrome/app/mash/mash_service_overrides.json --gtest_filter=....

Comment 5 by sky@chromium.org, Jan 13 2017

I generally look at the builder that runs mash browser_tests and copy the command line it uses.
Status: Started (was: Assigned)
Project Member

Comment 7 by bugdroid1@chromium.org, Feb 16 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/dcfe7f207b2b188b50b3e25eef5765b662c36455

commit dcfe7f207b2b188b50b3e25eef5765b662c36455
Author: jonross <jonross@chromium.org>
Date: Thu Feb 16 15:16:06 2017

Have WindowTreeClient detect error state earlier

Currently WindowTreeClient cleans up all windows during destruction. However
this masks incorrect usage of the client library. It also makes it difficult to
find errors, as crashes occur later on in shutdown.

This change updates the destructor to explicitly fail if there are windows.

TEST=manual, mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2696573003
Cr-Commit-Position: refs/heads/master@{#450968}

[modify] https://crrev.com/dcfe7f207b2b188b50b3e25eef5765b662c36455/ui/aura/mus/window_tree_client.cc
[modify] https://crrev.com/dcfe7f207b2b188b50b3e25eef5765b662c36455/ui/aura/mus/window_tree_client_observer.h
[modify] https://crrev.com/dcfe7f207b2b188b50b3e25eef5765b662c36455/ui/views/controls/native/native_view_host_aura_unittest.cc
[modify] https://crrev.com/dcfe7f207b2b188b50b3e25eef5765b662c36455/ui/views/mus/pointer_watcher_event_router.cc
[modify] https://crrev.com/dcfe7f207b2b188b50b3e25eef5765b662c36455/ui/views/mus/pointer_watcher_event_router.h

Comment 8 by sky@chromium.org, Mar 7 2017

Cc: roc...@chromium.org jamescook@chromium.org
 Issue 650311  has been merged into this issue.
Project Member

Comment 9 by bugdroid1@chromium.org, Mar 10 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/5530deaf657350413318fb3181f45ebe0f1954ba

commit 5530deaf657350413318fb3181f45ebe0f1954ba
Author: jonross <jonross@chromium.org>
Date: Fri Mar 10 22:37:06 2017

Update MojoTestConnection Service Registration

MojoTestConnection is responsible for creating the inter-process mojo pipes.
These must be set on the command line for the child processes running individual
test cases.

It is also responsible for bringing up services. This however was being done at
test enqueuing time. When the remote service was brought down at the end of the
first test case, all registered services were having their pipes close. Thus
preventing subsequent tests from connecting.

This change moves the service registration until the child process is being
connected. This actually occurs after the previous test has completed.

mash.browser_tests.filter has also been updated to include more tests which are
passing locally. This way we get expanded coverage on the bots.

TEST=BrowserTests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2737353003
Cr-Commit-Position: refs/heads/master@{#456192}

[modify] https://crrev.com/5530deaf657350413318fb3181f45ebe0f1954ba/chrome/test/base/mojo_test_connector.cc
[modify] https://crrev.com/5530deaf657350413318fb3181f45ebe0f1954ba/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 10 by bugdroid1@chromium.org, Mar 25 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/345a3681405a39f7ec3a675cc05457e04e708118

commit 345a3681405a39f7ec3a675cc05457e04e708118
Author: jonross <jonross@chromium.org>
Date: Sat Mar 25 02:17:32 2017

Enable more mash browser_tests

Update test filter to enable more tests.

TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2776453005
Cr-Commit-Position: refs/heads/master@{#459623}

[modify] https://crrev.com/345a3681405a39f7ec3a675cc05457e04e708118/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 11 by bugdroid1@chromium.org, Mar 27 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/466fa4cd13269cbd9159b13b61c6c35e1d75651c

commit 466fa4cd13269cbd9159b13b61c6c35e1d75651c
Author: jonross <jonross@chromium.org>
Date: Mon Mar 27 17:11:57 2017

Since the expansion of mash_browser_tests landed in: https://codereview.chromium.org/2776453005/

I'm expanding the Mojo FYI browser tests filter

TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2774353002
Cr-Commit-Position: refs/heads/master@{#459809}

[modify] https://crrev.com/466fa4cd13269cbd9159b13b61c6c35e1d75651c/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Blockedon: 706869
Blockedon: 706553

Comment 14 by sky@chromium.org, Apr 4 2017

Labels: mustash-2
Project Member

Comment 15 by bugdroid1@chromium.org, Apr 12 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/0190af327299091a9bdd376da2093cd48fe1f993

commit 0190af327299091a9bdd376da2093cd48fe1f993
Author: jonross <jonross@chromium.org>
Date: Wed Apr 12 14:18:21 2017

Based on running of the tests locally, along with on the trybots, the filter for mash_browser_tests have been updated with more tests

TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2786583003
Cr-Commit-Position: refs/heads/master@{#464015}

[modify] https://crrev.com/0190af327299091a9bdd376da2093cd48fe1f993/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 16 by bugdroid1@chromium.org, Apr 13 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/da791a0d74e8b12b9f758f6653cdc3299c55276b

commit da791a0d74e8b12b9f758f6653cdc3299c55276b
Author: tkent <tkent@chromium.org>
Date: Thu Apr 13 08:21:56 2017

Revert of Expand Mash Browser Test Coverage (patchset #11 id:220001 of https://codereview.chromium.org/2786583003/ )

Reason for revert:
Some tests are very flaky on ozone try bot.

https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_chromeos_ozone_rel_ng?numbuilds=200

Original issue's description:
> Based on running of the tests locally, along with on the trybots, the filter for mash_browser_tests have been updated with more tests
>
> TEST=mash_browser_tests
> BUG= 678687 
>
> Review-Url: https://codereview.chromium.org/2786583003
> Cr-Commit-Position: refs/heads/master@{#464015}
> Committed: https://chromium.googlesource.com/chromium/src/+/0190af327299091a9bdd376da2093cd48fe1f993

TBR=sky@chromium.org,jonross@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG= 678687 

Review-Url: https://codereview.chromium.org/2820483002
Cr-Commit-Position: refs/heads/master@{#464339}

[modify] https://crrev.com/da791a0d74e8b12b9f758f6653cdc3299c55276b/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 17 by bugdroid1@chromium.org, Apr 13 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/b62de227bd60c25f9a9b7421d628022fd2631a0c

commit b62de227bd60c25f9a9b7421d628022fd2631a0c
Author: jonross <jonross@chromium.org>
Date: Thu Apr 13 19:46:34 2017

More Mojo FYI Browser Tests

Recently the list of mash_browser_tests was expanded. However this uncovered
several flakes on the CQ.

This change update the FYI bot to test the list. So that we can identify if
there are still flakes before relanding on the main trybots.

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2812243004
Cr-Commit-Position: refs/heads/master@{#464512}

[modify] https://crrev.com/b62de227bd60c25f9a9b7421d628022fd2631a0c/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 18 by bugdroid1@chromium.org, Apr 18 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/4b7194fc64ee783975d60d364cb4b3cdeca58bd4

commit 4b7194fc64ee783975d60d364cb4b3cdeca58bd4
Author: jonross <jonross@chromium.org>
Date: Tue Apr 18 13:48:00 2017

Fix Mash Browser Test Startup Race

MojoTestConnector initialized a BackgroundServiceManager at creation time. Later
on it initializes mojo.

Due to this, it was possible for BackgroundServiceManager to attempt to create
mojo pipes for its catalog, before mojo has been initialized. Crashing the test
suite.

This change updates MojoTestConnector to delay the creation of
BackgroundServiceManager until it has explicitly created the mojo pipe.

TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2813293003
Cr-Commit-Position: refs/heads/master@{#465217}

[modify] https://crrev.com/4b7194fc64ee783975d60d364cb4b3cdeca58bd4/chrome/test/base/mojo_test_connector.cc
[modify] https://crrev.com/4b7194fc64ee783975d60d364cb4b3cdeca58bd4/chrome/test/base/mojo_test_connector.h

Project Member

Comment 19 by bugdroid1@chromium.org, Apr 18 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/00a08446f1e0bb1e6971a5f6d60a273ce78110d7

commit 00a08446f1e0bb1e6971a5f6d60a273ce78110d7
Author: jonross <jonross@chromium.org>
Date: Tue Apr 18 16:11:32 2017

Address Mojo FYI Browser Test Flakes

Based on the runs of the mash_browser_tests on the Mojo FYI bot this weekend two
flakes were found. This updates the filter to remove those.

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2827653002
Cr-Commit-Position: refs/heads/master@{#465256}

[modify] https://crrev.com/00a08446f1e0bb1e6971a5f6d60a273ce78110d7/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 20 by bugdroid1@chromium.org, Apr 21 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9eb30c04452a3b57e399417df86ae99ca03943ec

commit 9eb30c04452a3b57e399417df86ae99ca03943ec
Author: jonross <jonross@chromium.org>
Date: Fri Apr 21 15:30:42 2017

Cleanup Mojo FYI Filter

Sort and remove duplicates within the filter

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2831293003
Cr-Commit-Position: refs/heads/master@{#466343}

[modify] https://crrev.com/9eb30c04452a3b57e399417df86ae99ca03943ec/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 21 by bugdroid1@chromium.org, Apr 24 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/0ceb99da2db1300e2392a827203d791e426c1b9b

commit 0ceb99da2db1300e2392a827203d791e426c1b9b
Author: jonross <jonross@chromium.org>
Date: Mon Apr 24 17:05:08 2017

Update Mash Browser Test Filter

Based on results from the Mojo FYI bots running mash_browser_tests this is an
updated filter for the main trybots. All observed flakes have been removed.

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2830333002
Cr-Commit-Position: refs/heads/master@{#466665}

[modify] https://crrev.com/0ceb99da2db1300e2392a827203d791e426c1b9b/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 22 by bugdroid1@chromium.org, Apr 28 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/13d934bcc82c0e7ac86afe39fdede0a8d3b46d2a

commit 13d934bcc82c0e7ac86afe39fdede0a8d3b46d2a
Author: jonross <jonross@chromium.org>
Date: Fri Apr 28 19:53:43 2017

Update WindowManagerApplication teardown

WindowManagerApplication's normal destruction order involves tearing down
everything is has created.

However on some fast tests it is possible for WindowManagerApplication to be
destroyed before OnStart is ever called. When this occurs we attempt to destroy
services which we never created, and various DCHECKS are triggered.

PhishingDOMFeatureExtractorTest.SubframeRemoval has a >50% repro rate of this
locally where the crash occurs after the test successfully passes.

TEST=PhishingDOMFeatureExtractorTest.SubframeRemoval
BUG= 678687 

Review-Url: https://codereview.chromium.org/2851763002
Cr-Commit-Position: refs/heads/master@{#468101}

[modify] https://crrev.com/13d934bcc82c0e7ac86afe39fdede0a8d3b46d2a/ash/mus/window_manager_application.cc

Project Member

Comment 23 by bugdroid1@chromium.org, May 23 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d86012b80944319678d07adcc26b07b873deb5fa

commit d86012b80944319678d07adcc26b07b873deb5fa
Author: jonross <jonross@chromium.org>
Date: Tue May 23 15:06:01 2017

Have mash_browser_tests recreate BackgroundServiceManager per test

Currently one BackgroundServiceManager is created per batch of browser_tests.
This ends up leaving the UI and Ash process running between tests.

This is especially problematic for tests with PRE_test where state is expected
to not be preserved into the following test.

This change delays the initialization of a BackgroundServiceManager until the
child test process has connected to the MojoTestState. Subsequent tests will
tear down the previous BackgroundServiceManager before creating the new one.

This also tears down the UI and Ash processes as desired.

TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2858103002
Cr-Commit-Position: refs/heads/master@{#473897}

[modify] https://crrev.com/d86012b80944319678d07adcc26b07b873deb5fa/chrome/test/base/chrome_test_launcher.cc
[modify] https://crrev.com/d86012b80944319678d07adcc26b07b873deb5fa/chrome/test/base/mash_browser_tests_main.cc
[modify] https://crrev.com/d86012b80944319678d07adcc26b07b873deb5fa/chrome/test/base/mojo_test_connector.cc
[modify] https://crrev.com/d86012b80944319678d07adcc26b07b873deb5fa/chrome/test/base/mojo_test_connector.h
[modify] https://crrev.com/d86012b80944319678d07adcc26b07b873deb5fa/services/ui/service.cc
[modify] https://crrev.com/d86012b80944319678d07adcc26b07b873deb5fa/services/ui/service.h
[modify] https://crrev.com/d86012b80944319678d07adcc26b07b873deb5fa/testing/buildbot/chromium.chromiumos.json
[modify] https://crrev.com/d86012b80944319678d07adcc26b07b873deb5fa/testing/buildbot/filters/mash.browser_tests.filter

Blockedon: 725505
Project Member

Comment 25 by bugdroid1@chromium.org, May 25 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/189436107ee346b47eed5ed646ea9e5356a678fe

commit 189436107ee346b47eed5ed646ea9e5356a678fe
Author: jonross <jonross@chromium.org>
Date: Thu May 25 16:03:25 2017

Update mash_browser_tests FYI filter

There are a set of continually timing out tests on the FYI bots. Which are
blocking other tests from running.

I'm disabling them in the interim.

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2903153003
Cr-Commit-Position: refs/heads/master@{#474665}

[modify] https://crrev.com/189436107ee346b47eed5ed646ea9e5356a678fe/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 26 by bugdroid1@chromium.org, May 26 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/f45dbb0e42abd0c94f144ecb6966893a17759c07

commit f45dbb0e42abd0c94f144ecb6966893a17759c07
Author: jonross <jonross@chromium.org>
Date: Fri May 26 13:22:42 2017

One BackgroundServerManager per MojoTestState

We cannot have a shared BackgroundServerManager across all MojoTestStates.
Currently the shared BackgroundServerManager is restarted when each child test
process connects to the parent. However when batching tests they all connect at
the start, before running the tests in sequence.

This change makes it so that MojoTestState owns a BackgroundServerManager and
creates it when the child connects. Thus allowing each test case to have its
own set of the mash services running. Without other test runs deleting them.

When a child process ends the TestState is destroyed, which will cleanup the
services.

BUG= 678687 

Review-Url: https://codereview.chromium.org/2903533003
Cr-Commit-Position: refs/heads/master@{#474992}

[modify] https://crrev.com/f45dbb0e42abd0c94f144ecb6966893a17759c07/chrome/test/base/mash_browser_tests_main.cc
[modify] https://crrev.com/f45dbb0e42abd0c94f144ecb6966893a17759c07/chrome/test/base/mojo_test_connector.cc
[modify] https://crrev.com/f45dbb0e42abd0c94f144ecb6966893a17759c07/chrome/test/base/mojo_test_connector.h
[modify] https://crrev.com/f45dbb0e42abd0c94f144ecb6966893a17759c07/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 27 by bugdroid1@chromium.org, May 30 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/28f9bbec79ad97de51c4e99bb13359e1d691add0

commit 28f9bbec79ad97de51c4e99bb13359e1d691add0
Author: jonross <jonross@chromium.org>
Date: Tue May 30 17:55:27 2017

Disable new timeouts on Mojo FYI bot

The BrowserTest suite has begun timing out on the mash_browser_tests. Causing
the test suite to timeout before it can run all tests. Disabling while I
investigate the cause.

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2911233002
Cr-Commit-Position: refs/heads/master@{#475592}

[modify] https://crrev.com/28f9bbec79ad97de51c4e99bb13359e1d691add0/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 28 by bugdroid1@chromium.org, May 31 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c0e2d63185ee99b4262373458fb8937befe94b96

commit c0e2d63185ee99b4262373458fb8937befe94b96
Author: jonross <jonross@chromium.org>
Date: Wed May 31 15:21:03 2017

Update Mojo FYI bot for timeouts

Remove timeouts from filter list while we find out the cause.

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2915783002
Cr-Commit-Position: refs/heads/master@{#475913}

[modify] https://crrev.com/c0e2d63185ee99b4262373458fb8937befe94b96/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 29 by bugdroid1@chromium.org, Jun 1 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9f09049c45755baeee9ffd7e22c4e430ca96c2d4

commit 9f09049c45755baeee9ffd7e22c4e430ca96c2d4
Author: jonross <jonross@chromium.org>
Date: Thu Jun 01 15:55:32 2017

Update Mojo FYI filter

A few timeouts were missed. Removing them from the filter while the cause is
investigated offline.

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2913233002
Cr-Commit-Position: refs/heads/master@{#476311}

[modify] https://crrev.com/9f09049c45755baeee9ffd7e22c4e430ca96c2d4/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 30 by bugdroid1@chromium.org, Jun 23 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e4175877f5a72d240e89bde552f79fc31e2b0cad

commit e4175877f5a72d240e89bde552f79fc31e2b0cad
Author: jonross <jonross@chromium.org>
Date: Fri Jun 23 14:41:33 2017

Fix QuickLaunch AuraInit Crash

QuickLaunch will attempt to bring up its own Aura. However it is possible for
QuickLaunch::OnStart to occur after the peer has closed. AuraInit requires a
Directory api to load resources. Currently failing with a CHECK when it is not
available.

This updates AuraInit to report success/failure at initializing. QuickLaunch
uses this to cleanly close the ServiceContext on its side.

This will clean up a series of random crashes occuring in mash_browser_tests

TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2949103002
Cr-Commit-Position: refs/heads/master@{#481883}

[modify] https://crrev.com/e4175877f5a72d240e89bde552f79fc31e2b0cad/mash/quick_launch/quick_launch.cc
[modify] https://crrev.com/e4175877f5a72d240e89bde552f79fc31e2b0cad/ui/views/mus/aura_init.cc
[modify] https://crrev.com/e4175877f5a72d240e89bde552f79fc31e2b0cad/ui/views/mus/aura_init.h

I see mash_browser_tests often timing out on https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_chromeos_ozone_rel_ng.

When it doesn't time out, it comes very close to it, for example 14m 38s here:
https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_chromeos_ozone_rel_ng/builds/418589 (timeout is 15 minutes).

I suspect this is due to these DCHECKs firing:
[0629/150904.764785:FATAL:platform_font_linux.cc(62)] Check failed: typeface. Could not find any font: DejaVu Sans, sans
#0 0x000003750cfc base::debug::StackTrace::StackTrace()
#1 0x0000037699c1 logging::LogMessage::~LogMessage()
#2 0x000004a4bdd5 gfx::(anonymous namespace)::CreateSkTypeface()
#3 0x000004a4ba78 gfx::PlatformFontLinux::PlatformFontLinux()
#4 0x000004a4cdcb gfx::PlatformFont::CreateDefault()
#5 0x000004a3f4fe gfx::Font::Font()
#6 0x0000063b2625 views::AuraInit::AuraInit()
#7 0x000002bf965f _ZN4base10MakeUniqueIN5views8AuraInitEJPN15service_manager9ConnectorERKNS3_8IdentityERA22_KcRA26_S9_DnNS2_4ModeEEEENS_8internal16MakeUniqueResultIT_E6ScalarEDpOT0_
#8 0x000002bf9494 ash::mus::WindowManagerApplication::OnStart()
#9 0x0000041f66f6 service_manager::ServiceContext::OnStart()
#10 0x000004e92374 service_manager::mojom::ServiceStubDispatch::AcceptWithResponder()
#11 0x0000041f6ea6 service_manager::mojom::ServiceStub<>::AcceptWithResponder()
#12 0x000004dbc14d mojo::InterfaceEndpointClient::HandleValidatedMessage()
#13 0x000004dbbe16 mojo::FilterChain::Accept()
#14 0x000004dbd365 mojo::InterfaceEndpointClient::HandleIncomingMessage()
#15 0x000004dc431c mojo::internal::MultiplexRouter::ProcessIncomingMessage()
#16 0x000004dc3b0f mojo::internal::MultiplexRouter::Accept()
#17 0x000004dbbe16 mojo::FilterChain::Accept()
#18 0x000004dba179 mojo::Connector::ReadSingleMessage()
#19 0x000004dbaa31 mojo::Connector::ReadAllAvailableMessages()
#20 0x000004dba899 mojo::Connector::OnHandleReadyInternal()
#21 0x000004dd1689 mojo::SimpleWatcher::OnHandleReady()
#22 0x0000026876e3 _ZN4base8internal7InvokerINS0_9BindStateIMN7content25ServiceWorkerProviderHostEFviN5blink21WebServiceWorkerStateEEJNS_7WeakPtrIS4_EEiS6_EEEFvvEE3RunEPNS0_13BindStateBaseE
#23 0x0000006b2949 _ZNO4base8CallbackIFvvELNS_8internal8CopyModeE1ELNS2_10RepeatModeE1EE3RunEv
#24 0x000003803350 base::debug::TaskAnnotator::RunTask()
#25 0x000003771b3c base::MessageLoop::RunTask()
#26 0x000003771e4b base::MessageLoop::DeferOrRunPendingTask()
#27 0x000003772257 base::MessageLoop::DoWork()
#28 0x000003774879 base::MessagePumpLibevent::Run()
#29 0x0000037716fb base::MessageLoop::Run()
#30 0x00000379c24a base::RunLoop::Run()
#31 0x000003741992 (anonymous namespace)::StartEmbeddedService()
#32 0x0000037421cd _ZN4base8internal7InvokerINS0_9BindStateIPFvN4mojo16InterfaceRequestIN15service_manager5mojom7ServiceEEEEJEEES9_E3RunEPNS0_13BindStateBaseEOS8_
#33 0x00000210cedb service_manager::RunStandaloneService()
#34 0x00000374163e RunMashBrowserTests()
#35 0x0000037414a7 main
#36 0x7fc8399a0f45 __libc_start_main
#37 0x00000062360e <unknown>

Interesting that tests pass despite the crashes.
Also
[13089:13089:0629/152257.299005:FATAL:ozone_platform.cc(69)] Check failed: instance_. OzonePlatform is not initialized
#0 0x000003750cfc base::debug::StackTrace::StackTrace()
#1 0x0000037699c1 logging::LogMessage::~LogMessage()
#2 0x000001ed6182 ui::OzonePlatform::GetInstance()
#3 0x000006288ffe ash::ShellPortClassic::CreateNativeDisplayDelegate()
#4 0x0000062c430c ash::Shell::Init()
#5 0x0000062c3df9 ash::Shell::CreateInstance()
#6 0x0000067802fd AshInit::AshInit()
#7 0x000006580120 ChromeBrowserMainExtraPartsAsh::PreProfileInit()
#8 0x000003881d2a ChromeBrowserMainParts::PreProfileInit()
#9 0x000003882d80 ChromeBrowserMainPartsLinux::PreProfileInit()
#10 0x0000019755d3 chromeos::ChromeBrowserMainPartsChromeos::PreProfileInit()
#11 0x000003880faa ChromeBrowserMainParts::PreMainMessageLoopRunImpl()
#12 0x00000388092d ChromeBrowserMainParts::PreMainMessageLoopRun()
#13 0x000001975200 chromeos::ChromeBrowserMainPartsChromeos::PreMainMessageLoopRun()
#14 0x0000022e85f1 content::BrowserMainLoop::PreMainMessageLoopRun()
#15 0x0000026d5f07 content::StartupTaskRunner::RunAllTasksNow()
#16 0x0000022e6a7f content::BrowserMainLoop::CreateStartupTasks()
#17 0x0000022eaed2 content::BrowserMainRunnerImpl::Initialize()
#18 0x0000022e4097 content::BrowserMain()
#19 0x0000036fa03c content::ContentMainRunnerImpl::Run()
#20 0x00000564b064 service_manager::Main()
#21 0x0000036f8e84 content::ContentMain()
#22 0x000003df41e2 content::BrowserTestBase::SetUp()
#23 0x00000383fc33 InProcessBrowserTest::SetUp()
#24 0x000001c79991 testing::Test::Run()
#25 0x000001c7a4a0 testing::TestInfo::Run()
#26 0x000001c7a9d7 testing::TestCase::Run()
#27 0x000001c81b77 testing::internal::UnitTestImpl::RunAllTests()
#28 0x000001c817fb testing::UnitTest::Run()
#29 0x000003854066 base::TestSuite::Run()
#30 0x0000037438a3 ChromeTestSuiteRunner::RunTestSuite()
#31 0x000003e277d2 content::LaunchTests()
#32 0x0000037414f3 main
#33 0x7f1f4c09bf45 __libc_start_main
#34 0x00000062360e <unknown>
I also see some tests finishing with TIMED OUT status, like
[118/122] ExperimentalAppWindowApiTest.SetIcon (TIMED OUT)

Also, I see failing tests, even in successful runs:

5 tests failed:
    CreateNewFolder/FileManagerBrowserTest.Test/0 (../../chrome/browser/chromeos/file_manager/file_manager_browsertest.cc:40)
    CreateNewFolder/FileManagerBrowserTest.Test/1 (../../chrome/browser/chromeos/file_manager/file_manager_browsertest.cc:40)
    CreateNewFolder/FileManagerBrowserTest.Test/2 (../../chrome/browser/chromeos/file_manager/file_manager_browsertest.cc:40)
    CreateNewFolder/FileManagerBrowserTest.Test/3 (../../chrome/browser/chromeos/file_manager/file_manager_browsertest.cc:40)
    ExperimentalAppWindowApiTest.SetIcon (../../extensions/browser/api/app_window/app_window_apitest.cc:37)

Not sure what's going on here.
I've been looking at these crash stacks, as well as why we are close to the timeout.

I've filed two bugs to track them further
 Issue 738449  tracks the OzonePlatform crash as well as the length
 Issue 738441  tracks the AuraInit Font crash

The first issue does not trigger the test suite to fail, as it is occurring in a nested run of the test suite. The nesting of the test run is also what is pushing this to near the timeouts.

The second issue does not trigger crashes, as it is happening in a side process. The main process for the test case has actually already shut down at this time. But the side process was trying to startup, and is being scheduled then. It is trying to access resources which are no longer available.

So both types of crashes are happening outside of the scope of gtest HandleSehExceptionsInMethodIfSupported which is why they don't affect the actual results.
Project Member

Comment 35 by bugdroid1@chromium.org, Jul 6 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/14402447b7df16d16126dc49a3a44cf66c318c88

commit 14402447b7df16d16126dc49a3a44cf66c318c88
Author: jonross <jonross@chromium.org>
Date: Thu Jul 06 14:38:29 2017

It is possible for AuraInit to fail initialization. If the remote ServiceManager closes while a Service is starting up, then the synchronous calls to Directory will fail before error states are received.

This change follows up a previous change to AuraInit and QuickLauncher which was addressing these crashes in unittests. It refactors AuraInit to have an explicit initialization method. All callsites are updated to check the success of init, and to close cleanly upon failures.

TBR=sky@chromium.org
TEST=mash_browser_tests
BUG= 678687 

Review-Url: https://codereview.chromium.org/2967943002
Cr-Commit-Position: refs/heads/master@{#484583}

[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/ash/autoclick/mus/autoclick_application.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/ash/mus/window_manager_application.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/ash/touch_hud/mus/touch_hud_application.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/mash/catalog_viewer/catalog_viewer.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/mash/example/views_examples/views_examples.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/mash/example/window_type_launcher/window_type_launcher.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/mash/quick_launch/quick_launch.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/mash/simple_wm/simple_wm.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/mash/task_viewer/task_viewer.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/ui/views/mus/aura_init.cc
[modify] https://crrev.com/14402447b7df16d16126dc49a3a44cf66c318c88/ui/views/mus/aura_init.h

Project Member

Comment 36 by bugdroid1@chromium.org, Jul 6 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c

commit e3f9dbb7b50ab80caf29071bef886c2ec8cee28c
Author: fsamuel <fsamuel@chromium.org>
Date: Thu Jul 06 19:33:29 2017

Revert of Update AuraInit to handle failed initialization (patchset #2 id:40001 of https://codereview.chromium.org/2967943002/ )

Reason for revert:
Causing chrome --mash to fail to start

[118236:118236:0706/142624.101546:3509128143561:ERROR:chrome_main_delegate.cc(1187)] Main process exiting because service ash quit unexpectedly.

Original issue's description:
> It is possible for AuraInit to fail initialization. If the remote ServiceManager closes while a Service is starting up, then the synchronous calls to Directory will fail before error states are received.
>
> This change follows up a previous change to AuraInit and QuickLauncher which was addressing these crashes in unittests. It refactors AuraInit to have an explicit initialization method. All callsites are updated to check the success of init, and to close cleanly upon failures.
>
> TBR=sky@chromium.org
> TEST=mash_browser_tests
> BUG= 678687 
>
> Review-Url: https://codereview.chromium.org/2967943002
> Cr-Commit-Position: refs/heads/master@{#484583}
> Committed: https://chromium.googlesource.com/chromium/src/+/14402447b7df16d16126dc49a3a44cf66c318c88

TBR=sadrul@chromium.org,sky@chromium.org,jonross@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG= 678687 

Review-Url: https://codereview.chromium.org/2973103002
Cr-Commit-Position: refs/heads/master@{#484711}

[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/ash/autoclick/mus/autoclick_application.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/ash/mus/window_manager_application.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/ash/touch_hud/mus/touch_hud_application.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/mash/catalog_viewer/catalog_viewer.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/mash/example/views_examples/views_examples.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/mash/example/window_type_launcher/window_type_launcher.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/mash/quick_launch/quick_launch.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/mash/simple_wm/simple_wm.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/mash/task_viewer/task_viewer.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/ui/views/mus/aura_init.cc
[modify] https://crrev.com/e3f9dbb7b50ab80caf29071bef886c2ec8cee28c/ui/views/mus/aura_init.h

Project Member

Comment 37 by bugdroid1@chromium.org, Jul 10 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/1ab1bcf70718cc457e34d746b27ea2dd9a94871f

commit 1ab1bcf70718cc457e34d746b27ea2dd9a94871f
Author: Jonathan <jonross@chromium.org>
Date: Mon Jul 10 14:05:37 2017

Relnad Update AuraInit to handle failed initialization

This change relands AuraInit changes, with a fix to the shared resources logic
which was causing a crash on chrome --mash

Revert "Revert of Update AuraInit to handle failed initialization (patchset #2 id:40001 of https://codereview.chromium.org/2967943002/ )"

This reverts commit e3f9dbb7b50ab80caf29071bef886c2ec8cee28c.

TBR=sky@chromium.org
TEST=ran chrome --mash

Bug:  678687 
Change-Id: I6d64b04c520564a022b65e2e2f6c322803a9db64
Reviewed-on: https://chromium-review.googlesource.com/562497
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Reviewed-by: Sadrul Chowdhury <sadrul@chromium.org>
Cr-Commit-Position: refs/heads/master@{#485258}
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/ash/autoclick/mus/autoclick_application.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/ash/mus/window_manager_application.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/ash/touch_hud/mus/touch_hud_application.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/mash/catalog_viewer/catalog_viewer.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/mash/example/views_examples/views_examples.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/mash/example/window_type_launcher/window_type_launcher.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/mash/quick_launch/quick_launch.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/mash/simple_wm/simple_wm.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/mash/task_viewer/task_viewer.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/ui/views/mus/aura_init.cc
[modify] https://crrev.com/1ab1bcf70718cc457e34d746b27ea2dd9a94871f/ui/views/mus/aura_init.h

https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_chromeos_rel_ng?numbuilds=200

In 124 out of 200 attempts, we get "Exception steps exception mash_browser_tests (with patch) failed mash_browser_tests failed failure reason".

It fails with the following check in many cases:
[0711/025849.614516:FATAL:platform_font_linux.cc(174)] Check failed: success. Could not find any font: sans, sans
Thanks for letting me know that is occurring so much.

I landed a patch earlier which changed how we track font loading failures. While also fixing how the window manager handles the error state.

It appears that this has uncovered a flaky issue with ash::Shell::Init. That is being tracked in  issue 738441 
Project Member

Comment 40 by bugdroid1@chromium.org, Jul 28 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/eabe3b57cb3106eaaf3a7da698042fadfa162d1d

commit eabe3b57cb3106eaaf3a7da698042fadfa162d1d
Author: Jonathan <jonross@chromium.org>
Date: Fri Jul 28 14:20:33 2017

Fix Mojo FYI timeouts

Fix the timeout for mash_browser_tests/mus_browser_tests

Two problems here:
  - timeout is in seconds, not minutes
  - hard_timeout needs to be inside the swarming block, otherwise it's ignored

Right now the mash_browser_tests times out around 6mins, this should expand it
to be long enough to actually run the tests.

TBR=sky@chromium.org
TEST=none

Bug:  678687 
Change-Id: I3864936cc480831bd7d78d8a6375d42d850921c5
Reviewed-on: https://chromium-review.googlesource.com/589687
Reviewed-by: Jonathan Ross <jonross@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#490384}
[modify] https://crrev.com/eabe3b57cb3106eaaf3a7da698042fadfa162d1d/testing/buildbot/chromium.fyi.json

Project Member

Comment 41 by bugdroid1@chromium.org, Aug 1 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d11a643b986541db9ebe3e4a900e769596577105

commit d11a643b986541db9ebe3e4a900e769596577105
Author: Jonathan <jonross@chromium.org>
Date: Tue Aug 01 20:58:09 2017

Temporarily Incrase Mojo FYI timeouts

Even with properly set timeouts, mash_browser_tests timeout about half the time
on the Mojo FYI bot. I want to see if an increased timeout can get them
consistently passing.

TBR=sky@chromium.org
TEST=none

Bug:  678687 
Change-Id: I1b86980f981a3c00e5caf1c4aaae9f4c6ad06d8f
Reviewed-on: https://chromium-review.googlesource.com/593851
Reviewed-by: Jonathan Ross <jonross@chromium.org>
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#491106}
[modify] https://crrev.com/d11a643b986541db9ebe3e4a900e769596577105/testing/buildbot/chromium.fyi.json

Project Member

Comment 42 by bugdroid1@chromium.org, Aug 4 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9d3c07877e454d4540255b48f458b0c0d59f16fc

commit 9d3c07877e454d4540255b48f458b0c0d59f16fc
Author: Jonathan <jonross@chromium.org>
Date: Fri Aug 04 16:42:30 2017

Disable CreateNewFolder tests on Mojo FYI

In the past there were observed flakes of CreateNewFolder/FileManagerBrowserTest
tracked in  issue 715963 . These flakes disappeared and the tests were re-enabled.

They appear to be flaking on the mash_browser_tests, so I'm disabling them on
the Mojo FYI bot for now.

TBR=sky@chromium.org
TEST=mash_browser_tests

Bug:  678687 
Change-Id: Iea296a31b8ab80a76edd17e1d33de20fd5717d17
Reviewed-on: https://chromium-review.googlesource.com/602058
Reviewed-by: Jonathan Ross <jonross@chromium.org>
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#492050}
[modify] https://crrev.com/9d3c07877e454d4540255b48f458b0c0d59f16fc/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 43 by bugdroid1@chromium.org, Aug 9 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/5bb9470c491da07b113361302658008bbca07ac9

commit 5bb9470c491da07b113361302658008bbca07ac9
Author: Jonathan <jonross@chromium.org>
Date: Wed Aug 09 06:47:25 2017

Update Mash Browser Test Sharding

Update mash_browser_tests on the Mojo FYI bots to shard. Hopefully we can better
track down what is causing the full suite to timeout. It's a flake that is not
reproducing locally

TBR=sky@chromium.org
TEST=mash_browser_tests

Bug:  678687 
Change-Id: I09a4d8210fe34f44c9e8fa4bc2f4e0a49abf2261
Reviewed-on: https://chromium-review.googlesource.com/606567
Reviewed-by: Jonathan Ross <jonross@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#492864}
[modify] https://crrev.com/5bb9470c491da07b113361302658008bbca07ac9/testing/buildbot/chromium.fyi.json

Project Member

Comment 44 by bugdroid1@chromium.org, Aug 9 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/18963aa59a11a3001b8dda5b3a38a81e13480533

commit 18963aa59a11a3001b8dda5b3a38a81e13480533
Author: Jonathan <jonross@chromium.org>
Date: Wed Aug 09 17:33:36 2017

Disable Suspect Test in Mojo FYI mash_browser_tests

With the mash_browser_tests sharding we can narrow in on tests that are causing
flaky timeouts. In one run a shard only ran a single test case and timed out.
This test was also seen in each other shard that was timing out.

FormAutocompleteTest.AjaxSucceeded_StillVisible is being disabled on the Mojo
FYI mash_browser_tests bots.

TBR=sky@chromium.org
TEST=mash_browser_tests

Bug:  678687 
Change-Id: I2c9af0acbc8e9ed6962b93e804224ea749216daa
Reviewed-on: https://chromium-review.googlesource.com/608614
Reviewed-by: Jonathan Ross <jonross@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#493042}
[modify] https://crrev.com/18963aa59a11a3001b8dda5b3a38a81e13480533/testing/buildbot/filters/mojo.fyi.browser_tests.filter

Project Member

Comment 45 by bugdroid1@chromium.org, Aug 19 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d705f0e937421361bd03a8b0382807001a1b753e

commit d705f0e937421361bd03a8b0382807001a1b753e
Author: Jonathan <jonross@chromium.org>
Date: Sat Aug 19 00:37:12 2017

Setup a single test version of mash_browser_test

With a deadlock found in the full mash_browser_tests suite, we are wondering if
a single test can be ran reliably.

This adds a new config to Mojo FYI bot for CrOS to verify this._

TEST=mash_browser_tests

Bug:  678687 
Change-Id: I374e36a6fb23d27c95c2b058ca7706f2466c115f
Reviewed-on: https://chromium-review.googlesource.com/621230
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#495765}
[modify] https://crrev.com/d705f0e937421361bd03a8b0382807001a1b753e/testing/buildbot/chromium.fyi.json
[modify] https://crrev.com/d705f0e937421361bd03a8b0382807001a1b753e/testing/buildbot/filters/BUILD.gn
[add] https://crrev.com/d705f0e937421361bd03a8b0382807001a1b753e/testing/buildbot/filters/mojo.fyi.single.mash.browser_tests.filter

Blockedon: 757511
Blockedon: 760257
Blockedon: 760298
Blockedon: 735051

Comment 50 by sky@chromium.org, Sep 8 2017

Blockedon: 763529
I'm adding 763529 to the list of blocked bugs. Fixing 763529 may change the nature of the deadlocks we're seeing.
Project Member

Comment 51 by bugdroid1@chromium.org, Nov 7 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/14b4134077b54e0278bb651e008fd069a6c539a7

commit 14b4134077b54e0278bb651e008fd069a6c539a7
Author: James Cook <jamescook@chromium.org>
Date: Tue Nov 07 21:03:14 2017

cros: Adapt ShelfBrowserTest to work with --mash

* Add a mojo shelf test API
* Convert 2 tests to use it

Bug: 781925,  678687 
Test: browser_tests --mash
Change-Id: If75111a9795bc93eefb69bc2d80401d1fb2b6d65
Reviewed-on: https://chromium-review.googlesource.com/755344
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#514582}
[modify] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/ash/BUILD.gn
[modify] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/ash/mojo_test_interface_factory.cc
[modify] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/ash/mus/manifest.json
[modify] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/ash/public/interfaces/BUILD.gn
[add] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/ash/public/interfaces/shelf_test_api.mojom
[add] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/ash/shelf/shelf_test_api.cc
[add] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/ash/shelf/shelf_test_api.h
[modify] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/chrome/browser/ui/ash/shelf_browsertest.cc
[modify] https://crrev.com/14b4134077b54e0278bb651e008fd069a6c539a7/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 52 by bugdroid1@chromium.org, Nov 7 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c11e7ba721840d965f6ae24e7f72ab260d166f5e

commit c11e7ba721840d965f6ae24e7f72ab260d166f5e
Author: James Cook <jamescook@chromium.org>
Date: Tue Nov 07 22:30:01 2017

cros: Replace shelf STATUS_ACTIVE with STATUS_RUNNING

Before the material design shelf we used to show light bars underneath
the shelf icons for running apps. An app with an active window had a
brighter bar. With the MD redesign we show dots and there is no
difference between running and active. Tracking the active state
complicates the go/mustash refactor, especially in tests.

* Eliminate the active state, which simplifies both code and tests
* Remove ShelfButton::STATE_ACTIVE and STATE_FOCUSED, neither of
  which are used anymore
* Eliminate browser tests that existed mostly to track the active
  state

Bug:  678687 
Test: ash_unittests, browser_tests
Change-Id: I3dc1b045d44ea1ecc5a311afa4d877d0e908e35a
Reviewed-on: https://chromium-review.googlesource.com/757169
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#514623}
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/public/cpp/shelf_struct_traits.h
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/public/cpp/shelf_types.h
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/public/interfaces/shelf.mojom
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/shelf/shelf_button.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/shelf/shelf_button.h
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/shelf/shelf_view.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/shelf/shelf_view_unittest.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/shelf/shelf_window_watcher.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/ash/shelf/shelf_window_watcher_unittest.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/app_window_launcher_controller.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/app_window_launcher_item_controller.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/app_window_launcher_item_controller.h
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/arc_app_window_launcher_controller.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/browser_shortcut_launcher_item_controller.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/chrome_launcher_controller.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/chrome_launcher_controller.h
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/chrome_launcher_controller_browsertest.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/chrome_launcher_controller_unittest.cc
[modify] https://crrev.com/c11e7ba721840d965f6ae24e7f72ab260d166f5e/chrome/browser/ui/ash/launcher/extension_app_window_launcher_controller.cc

Project Member

Comment 53 by bugdroid1@chromium.org, Nov 8 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/347637a52653723220d64c10b567d22b7c0074e8

commit 347637a52653723220d64c10b567d22b7c0074e8
Author: Scott Violet <sky@chromium.org>
Date: Wed Nov 08 00:55:02 2017

Revert "cros: Replace shelf STATUS_ACTIVE with STATUS_RUNNING"

This reverts commit c11e7ba721840d965f6ae24e7f72ab260d166f5e.

Reason for revert: Appears to have caused mash_browsertests to fail. https://chromium-swarm.appspot.com/task?id=39b1d0d34a750d10&refresh=10&show_raw=1&wide_logs=true output is:

../../chrome/browser/ui/ash/shelf_browsertest.cc:76: Failure
Value of: has_overlapping_window
  Actual: false
Expected: true
[500:500:1107/162023.401861:INFO:chrome_cryptauth_service.cc(230)] Refresh token not yet available; waiting before starting CryptAuth managers.
[534:534:1107/162023.407440:ERROR:shell_delegate_mus.cc(75)] Not implemented reached in virtual void ash::ShellDelegateMus::PreShutdown()
[  FAILED  ] ShelfBrowserTest.StatusBubble, where TypeParam =  and GetParam() =  (962 ms)


Original change's description:
> cros: Replace shelf STATUS_ACTIVE with STATUS_RUNNING
> 
> Before the material design shelf we used to show light bars underneath
> the shelf icons for running apps. An app with an active window had a
> brighter bar. With the MD redesign we show dots and there is no
> difference between running and active. Tracking the active state
> complicates the go/mustash refactor, especially in tests.
> 
> * Eliminate the active state, which simplifies both code and tests
> * Remove ShelfButton::STATE_ACTIVE and STATE_FOCUSED, neither of
>   which are used anymore
> * Eliminate browser tests that existed mostly to track the active
>   state
> 
> Bug:  678687 
> Test: ash_unittests, browser_tests
> Change-Id: I3dc1b045d44ea1ecc5a311afa4d877d0e908e35a
> Reviewed-on: https://chromium-review.googlesource.com/757169
> Reviewed-by: Tom Sepez <tsepez@chromium.org>
> Reviewed-by: Michael Wasserman <msw@chromium.org>
> Commit-Queue: James Cook <jamescook@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#514623}

TBR=jamescook@chromium.org,msw@chromium.org,tsepez@chromium.org

Change-Id: Ibf53d35870f583c86519b6b18b2036d64c82c1f7
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug:  678687 
Reviewed-on: https://chromium-review.googlesource.com/757937
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#514685}
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/public/cpp/shelf_struct_traits.h
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/public/cpp/shelf_types.h
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/public/interfaces/shelf.mojom
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/shelf/shelf_button.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/shelf/shelf_button.h
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/shelf/shelf_view.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/shelf/shelf_view_unittest.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/shelf/shelf_window_watcher.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/ash/shelf/shelf_window_watcher_unittest.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/app_window_launcher_controller.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/app_window_launcher_item_controller.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/app_window_launcher_item_controller.h
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/arc_app_window_launcher_controller.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/browser_shortcut_launcher_item_controller.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/chrome_launcher_controller.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/chrome_launcher_controller.h
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/chrome_launcher_controller_browsertest.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/chrome_launcher_controller_unittest.cc
[modify] https://crrev.com/347637a52653723220d64c10b567d22b7c0074e8/chrome/browser/ui/ash/launcher/extension_app_window_launcher_controller.cc

Project Member

Comment 54 by bugdroid1@chromium.org, Nov 8 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9716183168af9fd41917aca3ce613b9801b670c3

commit 9716183168af9fd41917aca3ce613b9801b670c3
Author: James Cook <jamescook@chromium.org>
Date: Wed Nov 08 02:56:57 2017

Revert "cros: Adapt ShelfBrowserTest to work with --mash"

This reverts commit 14b4134077b54e0278bb651e008fd069a6c539a7.

Reason for revert: Causes flaky failures in mash_browser_tests. I can repro the failures locally. Sample from bots:

https://chromium-swarm.appspot.com/task?id=39b1d0d34a750d10&refresh=10&show_raw=1&wide_logs=true output is:
    
    ../../chrome/browser/ui/ash/shelf_browsertest.cc:76: Failure
    Value of: has_overlapping_window
      Actual: false
    Expected: true
    [500:500:1107/162023.401861:INFO:chrome_cryptauth_service.cc(230)] Refresh token not yet available; waiting before starting CryptAuth managers.
    [534:534:1107/162023.407440:ERROR:shell_delegate_mus.cc(75)] Not implemented reached in virtual void ash::ShellDelegateMus::PreShutdown()
    [  FAILED  ] ShelfBrowserTest.StatusBubble, where TypeParam =  and GetParam() =  (962 ms)

Original change's description:
> cros: Adapt ShelfBrowserTest to work with --mash
> 
> * Add a mojo shelf test API
> * Convert 2 tests to use it
> 
> Bug: 781925,  678687 
> Test: browser_tests --mash
> Change-Id: If75111a9795bc93eefb69bc2d80401d1fb2b6d65
> Reviewed-on: https://chromium-review.googlesource.com/755344
> Reviewed-by: Tom Sepez <tsepez@chromium.org>
> Reviewed-by: Michael Wasserman <msw@chromium.org>
> Commit-Queue: James Cook <jamescook@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#514582}

TBR=jamescook@chromium.org,msw@chromium.org,tsepez@chromium.org

Change-Id: I6e4bea721b44cacb2cf53d9c8861e7d4889fb4de
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 781925,  678687 
Reviewed-on: https://chromium-review.googlesource.com/757792
Reviewed-by: James Cook <jamescook@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#514722}
[modify] https://crrev.com/9716183168af9fd41917aca3ce613b9801b670c3/ash/BUILD.gn
[modify] https://crrev.com/9716183168af9fd41917aca3ce613b9801b670c3/ash/mojo_test_interface_factory.cc
[modify] https://crrev.com/9716183168af9fd41917aca3ce613b9801b670c3/ash/mus/manifest.json
[modify] https://crrev.com/9716183168af9fd41917aca3ce613b9801b670c3/ash/public/interfaces/BUILD.gn
[delete] https://crrev.com/e704e6785d9a47fe89e5030b9a3d7340360d5709/ash/public/interfaces/shelf_test_api.mojom
[delete] https://crrev.com/e704e6785d9a47fe89e5030b9a3d7340360d5709/ash/shelf/shelf_test_api.cc
[delete] https://crrev.com/e704e6785d9a47fe89e5030b9a3d7340360d5709/ash/shelf/shelf_test_api.h
[modify] https://crrev.com/9716183168af9fd41917aca3ce613b9801b670c3/chrome/browser/ui/ash/shelf_browsertest.cc
[modify] https://crrev.com/9716183168af9fd41917aca3ce613b9801b670c3/testing/buildbot/filters/mash.browser_tests.filter

Components: Internals>Services>ServiceManager
Bulk applying component Internals>Services>ServiceManager to issues referencing the text ServiceManager.  This may not be 100% accurate, so please feel free to pull the component as needed.
Project Member

Comment 56 by bugdroid1@chromium.org, Nov 8 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/263e93e3e6ad65dc66abad514c3cf3e4c1010af8

commit 263e93e3e6ad65dc66abad514c3cf3e4c1010af8
Author: James Cook <jamescook@chromium.org>
Date: Wed Nov 08 17:37:24 2017

Reland: cros: Replace shelf STATUS_ACTIVE with STATUS_RUNNING

This was reverted due to failures in mash_browser_tests for
ShelfBrowserTest.StatusBubble. The problem was actually
due to 9716183168af9fd41917aca3ce613b9801b670c3 and that CL has been
reverted.

    > cros: Replace shelf STATUS_ACTIVE with STATUS_RUNNING
    >
    > Before the material design shelf we used to show light bars underneath
    > the shelf icons for running apps. An app with an active window had a
    > brighter bar. With the MD redesign we show dots and there is no
    > difference between running and active. Tracking the active state
    > complicates the go/mustash refactor, especially in tests.
    >
    > * Eliminate the active state, which simplifies both code and tests
    > * Remove ShelfButton::STATE_ACTIVE and STATE_FOCUSED, neither of
    >   which are used anymore
    > * Eliminate browser tests that existed mostly to track the active
    >   state
    >
    > Bug:  678687 
    > Test: ash_unittests, browser_tests
    > Change-Id: I3dc1b045d44ea1ecc5a311afa4d877d0e908e35a
    > Reviewed-on: https://chromium-review.googlesource.com/757169
    > Reviewed-by: Tom Sepez <tsepez@chromium.org>
    > Reviewed-by: Michael Wasserman <msw@chromium.org>
    > Commit-Queue: James Cook <jamescook@chromium.org>
    > Cr-Commit-Position: refs/heads/master@{#514623}

    TBR=jamescook@chromium.org,msw@chromium.org,tsepez@chromium.org
    Bug:  678687 
    Reviewed-on: https://chromium-review.googlesource.com/757937
    Reviewed-by: Scott Violet <sky@chromium.org>
    Commit-Queue: Scott Violet <sky@chromium.org>
    Cr-Commit-Position: refs/heads/master@{#514685}

Change-Id: Ief70c56e406ed27394716ea0752d10530aefb1f5
Reviewed-on: https://chromium-review.googlesource.com/758800
Reviewed-by: James Cook <jamescook@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#514872}
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/public/cpp/shelf_struct_traits.h
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/public/cpp/shelf_types.h
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/public/interfaces/shelf.mojom
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/shelf/shelf_button.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/shelf/shelf_button.h
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/shelf/shelf_view.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/shelf/shelf_view_unittest.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/shelf/shelf_window_watcher.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/ash/shelf/shelf_window_watcher_unittest.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/app_window_launcher_controller.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/app_window_launcher_item_controller.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/app_window_launcher_item_controller.h
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/arc_app_window_launcher_controller.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/browser_shortcut_launcher_item_controller.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/chrome_launcher_controller.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/chrome_launcher_controller.h
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/chrome_launcher_controller_browsertest.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/chrome_launcher_controller_unittest.cc
[modify] https://crrev.com/263e93e3e6ad65dc66abad514c3cf3e4c1010af8/chrome/browser/ui/ash/launcher/extension_app_window_launcher_controller.cc

Project Member

Comment 57 by bugdroid1@chromium.org, Nov 9 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/8f09721fadf98c35df07c2db04634135a7c72ea2

commit 8f09721fadf98c35df07c2db04634135a7c72ea2
Author: James Cook <jamescook@chromium.org>
Date: Thu Nov 09 00:18:54 2017

Reland: cros: Adapt ShelfBrowserTest to work with --mash

The original CL caused mash_browser_test flake because the test was not
waiting for in-flight window bounds/visibility changes to apply, so
there was a race between browser, window server and ash.

Original CL description:
* Add a mojo shelf test API
* Convert 2 tests to use it

Bug: 781925,  678687 
Test: browser_tests --mash
Change-Id: Icfe23d4c8c3bcd84bff8dc46d2c9680cc656ce50
Reviewed-on: https://chromium-review.googlesource.com/755344
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#514582}

TBR=tsepez@chromium.org

Change-Id: Icfe23d4c8c3bcd84bff8dc46d2c9680cc656ce50
Reviewed-on: https://chromium-review.googlesource.com/759034
Reviewed-by: James Cook <jamescook@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515028}
[modify] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/ash/BUILD.gn
[modify] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/ash/mojo_test_interface_factory.cc
[modify] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/ash/mus/manifest.json
[modify] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/ash/public/interfaces/BUILD.gn
[add] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/ash/public/interfaces/shelf_test_api.mojom
[add] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/ash/shelf/shelf_test_api.cc
[add] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/ash/shelf/shelf_test_api.h
[modify] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/chrome/browser/ui/ash/shelf_browsertest.cc
[modify] https://crrev.com/8f09721fadf98c35df07c2db04634135a7c72ea2/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 58 by bugdroid1@chromium.org, Nov 10 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/514d6492a599d544e46b2508c54f69a4b7bf16ea

commit 514d6492a599d544e46b2508c54f69a4b7bf16ea
Author: James Cook <jamescook@chromium.org>
Date: Thu Nov 09 23:59:33 2017

cros: Add extension tests to mash browser_tests on mojo FYI bot

We'll see how stable they are over the weekend.

Bug:  678687 
Change-Id: I85c50f47f1e1f148f7e52bc7f09be7cf5eee4d25

TBR=sky@chromium.org

Change-Id: I85c50f47f1e1f148f7e52bc7f09be7cf5eee4d25
Reviewed-on: https://chromium-review.googlesource.com/762403
Reviewed-by: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#515350}
[modify] https://crrev.com/514d6492a599d544e46b2508c54f69a4b7bf16ea/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

@ynovikov - This bug is for mash_browser_tests. The failed build you linked is for mus_browser_tests. I see some mus_browser_tests failures here, but most of them are on builds where regular browser_tests fail, so I would expect mus_browser_tests to fail as well:

https://ci.chromium.org/buildbot/tryserver.chromium.linux/linux_chromium_chromeos_rel_ng/

Have you been seeing cases where mash_browser_tests is flaking? Or where mus_browser_tests is failing but regular browser_tests are not?

Ah, I'm sorry, indeed, I've confused mus_browser_tests and mash_browser_tests.
Probably because #40 refers to both.
Project Member

Comment 62 by bugdroid1@chromium.org, Nov 14 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2eed98790035408faf283311630862127d24d36f

commit 2eed98790035408faf283311630862127d24d36f
Author: James Cook <jamescook@chromium.org>
Date: Tue Nov 14 21:08:54 2017

cros: Fix some shelf browser tests under --mash

* Don't start quick launch in tests, because that changes
  the number of shelf items vs. classic ash
* Fix "draw attention" property mirroring, such that
  ash updates when chrome sets the property
* Eliminate direct access to ash::Shell and ash::Shelf*
  in ChromeLauncherController test setup, which fixes a
  bunch of test crashes

Bug:  678687 
Test: browser_tests --mash
Change-Id: I721d95b42f2bf36fffbfde94db32db46626a48df
Reviewed-on: https://chromium-review.googlesource.com/767168
Commit-Queue: James Cook <jamescook@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#516424}
[modify] https://crrev.com/2eed98790035408faf283311630862127d24d36f/ash/mus/window_manager.cc
[modify] https://crrev.com/2eed98790035408faf283311630862127d24d36f/chrome/browser/ui/ash/launcher/chrome_launcher_controller_browsertest.cc
[modify] https://crrev.com/2eed98790035408faf283311630862127d24d36f/chrome/browser/ui/views/chrome_browser_main_extra_parts_views.cc
[modify] https://crrev.com/2eed98790035408faf283311630862127d24d36f/services/ui/public/interfaces/window_manager.mojom
[modify] https://crrev.com/2eed98790035408faf283311630862127d24d36f/ui/aura/mus/property_converter.cc

Project Member

Comment 63 by bugdroid1@chromium.org, Nov 14 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/98410e31657cc039987ce21c7481178842a2d94b

commit 98410e31657cc039987ce21c7481178842a2d94b
Author: James Cook <jamescook@chromium.org>
Date: Tue Nov 14 22:00:50 2017

cros: Fix dialog browser_tests under --mash

TestBrowserDialog collects a list of all top-level widgets to ensure
that a single dialog is opened by each test. On Chrome OS this directly
accesses ash::Shell to get the primary root window and build a list of
children. This doesn't work under mash and isn't needed anyhow, since
views::test::WidgetTest uses MusClient to get a list of root windows
under mash.

This fixes ~100 crashing tests under --mash.

Bug:  678687 
Test: browser_tests --mash
Change-Id: I2c3ac797826e6cd1b8fc747955a86a926912f60c
Reviewed-on: https://chromium-review.googlesource.com/769589
Commit-Queue: James Cook <jamescook@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#516447}
[modify] https://crrev.com/98410e31657cc039987ce21c7481178842a2d94b/chrome/browser/ui/test/test_browser_dialog.cc
[modify] https://crrev.com/98410e31657cc039987ce21c7481178842a2d94b/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Project Member

Comment 64 by bugdroid1@chromium.org, Nov 15 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/aa35905a3b1aafccb6ca9f964b4bcd00bfd92f8c

commit aa35905a3b1aafccb6ca9f964b4bcd00bfd92f8c
Author: James Cook <jamescook@chromium.org>
Date: Wed Nov 15 15:55:51 2017

cros: Fix use-after-free in fullscreen window teardown

ImmersiveModeControllerAsh caches a pointer to the BrowserView's native
aura::Window, which it used to clean up observers when the browser
window was closed. However, Browser/BrowserView teardown deletes that
aura::Window first, so this was a use-after-free whenever a window
was closed while still in the fullscreen state.

Stop caching the pointer and clean up observers on aura::Window
destruction instead.

This also fixes BrowserTest.FullscreenBookmarkBar under --mash.

Bug:  678687 
Test: browser_tests BrowserTest.* with and without --mash
Change-Id: Ib1af8732d28d277dff68ab2dfda8d728f78d2213
Reviewed-on: https://chromium-review.googlesource.com/770313
Reviewed-by: Scott Violet <sky@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#516707}
[modify] https://crrev.com/aa35905a3b1aafccb6ca9f964b4bcd00bfd92f8c/chrome/browser/ui/browser_browsertest.cc
[modify] https://crrev.com/aa35905a3b1aafccb6ca9f964b4bcd00bfd92f8c/chrome/browser/ui/views/frame/immersive_mode_controller_ash.cc
[modify] https://crrev.com/aa35905a3b1aafccb6ca9f964b4bcd00bfd92f8c/chrome/browser/ui/views/frame/immersive_mode_controller_ash.h
[modify] https://crrev.com/aa35905a3b1aafccb6ca9f964b4bcd00bfd92f8c/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Project Member

Comment 65 by bugdroid1@chromium.org, Nov 17 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/bd0b779e15d7dc045e4d38f3a02a3445457aaafd

commit bd0b779e15d7dc045e4d38f3a02a3445457aaafd
Author: James Cook <jamescook@chromium.org>
Date: Fri Nov 17 03:24:26 2017

cros: Move touch HUD enabled preference from chrome to ash

This eliminates some direct calls from browser into ash, which we need
to do for go/mustash (out-of-process ash).

Introduce TouchHudController to observe multi-user changes and update
the per-display configuration.

Add unit test coverage in ash and eliminate browser test calls. This
is not yet enough to enable PreferencesTest.MultiProfiles with --mash
(due to IME issues) but it's a step forward.

Bug:  678687 
Test: added to ash_unittests, manually use Ctrl-Alt-P to enable touch hud with multiple users and multiple displays

Change-Id: Id86679a20af421a9e6532a79658e83d0d21b473c
Reviewed-on: https://chromium-review.googlesource.com/772839
Commit-Queue: James Cook <jamescook@chromium.org>
Reviewed-by: Michael Wasserman <msw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#517276}
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/accelerators/accelerator_commands_classic.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/public/cpp/ash_pref_names.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/public/cpp/ash_pref_names.h
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/root_window_controller.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/root_window_controller.h
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/session/session_observer.h
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/shell.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/shell.h
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/shell_observer.h
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/touch/touch_devices_controller.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/touch/touch_devices_controller.h
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/touch/touch_devices_controller_unittest.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/ash/touch/touch_observer_hud_unittest.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/chrome/browser/chromeos/preferences.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/chrome/browser/chromeos/preferences.h
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/chrome/browser/chromeos/preferences_chromeos_browsertest.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/chrome/common/pref_names.cc
[modify] https://crrev.com/bd0b779e15d7dc045e4d38f3a02a3445457aaafd/chrome/common/pref_names.h

Project Member

Comment 66 by bugdroid1@chromium.org, Nov 17 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/218fb925ecd32f424842c68f6ba14a660de2b0a3

commit 218fb925ecd32f424842c68f6ba14a660de2b0a3
Author: James Cook <jamescook@chromium.org>
Date: Fri Nov 17 15:24:52 2017

chromeos: Categorize some browser_tests --mash failures in filter file

This is a start toward grouping them by cause. Also enable these tests,
which pass now:

AllForms*
AllProcesses*
BlockingLoginTestInstance/BlockingLoginTest.*
BrailleDisplayPrivateAPIUserTest.*
BrowserCloseManagerBrowserTest/BrowserCloseManagerBrowserTest.*
BrowserCloseManagerWithDownloadsBrowserTest/BrowserCloseManagerWithDownloadsBrowserTest.*
BrowserLoginTest.*
CertificateReportingServiceBrowserTest.*
ChromeOriginTrialsDisabledTokensTest.*
ContentScriptApiTests/ContentScriptApiTest.*
CrControlledButtonTest.*
CrControlledRadioButtonTest.*
CrExtensionsManagerTestWithMultipleExtensionTypesInstalled.*

Bug:  678687 
Change-Id: I6eae99411a5ac0ce72d098da06ca7e42399181cd
Reviewed-on: https://chromium-review.googlesource.com/776075
Reviewed-by: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#517393}
[modify] https://crrev.com/218fb925ecd32f424842c68f6ba14a660de2b0a3/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Cc: -jamescook@chromium.org steve...@chromium.org
Components: -Internals>Services>ServiceManager Internals>MUS
Labels: -mustash-1 -mustash-2
Owner: jamescook@chromium.org
I can own this for now.

Project Member

Comment 68 by bugdroid1@chromium.org, Nov 17 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d432e9f2601140f2ceef14a3d5ee0af03bf248aa

commit d432e9f2601140f2ceef14a3d5ee0af03bf248aa
Author: James Cook <jamescook@chromium.org>
Date: Fri Nov 17 19:42:45 2017

cros: Update FYI bot mash_browser_tests filter file for D,E,F

Document test failures and re-enable passing tests. This brings us to
5200 tests.

Bug:  678687 
Change-Id: I800f19093e4e614e69fa6d311d9bb767c9e3e759
Reviewed-on: https://chromium-review.googlesource.com/777546
Reviewed-by: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#517516}
[modify] https://crrev.com/d432e9f2601140f2ceef14a3d5ee0af03bf248aa/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Project Member

Comment 69 by bugdroid1@chromium.org, Nov 18 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/b7be1392a7f489bad83b0ee3183b7fcc63b222a6

commit b7be1392a7f489bad83b0ee3183b7fcc63b222a6
Author: James Cook <jamescook@chromium.org>
Date: Sat Nov 18 01:20:41 2017

cros: Update mojo.fyi.mash.browser_tests.filter file

Enable all passing tests. Categorize all failing tests and file bugs
for the most common ones.

Bug:  678687 
Test: browser_tests --mash
Change-Id: I13bbf84322aaa3d6dcceb36a2fceff1fc16280f1
Reviewed-on: https://chromium-review.googlesource.com/777848
Reviewed-by: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#517663}
[modify] https://crrev.com/b7be1392a7f489bad83b0ee3183b7fcc63b222a6/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Blockedon: 787121
Project Member

Comment 71 by bugdroid1@chromium.org, Nov 20 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/49e43eff449b72d330b32152accbb023d4622879

commit 49e43eff449b72d330b32152accbb023d4622879
Author: Jonathan <jonross@chromium.org>
Date: Mon Nov 20 23:25:29 2017

Disable flaking PreferencesTest.MultiProfiles

Currently PreferencesTest.MultiProfiles is flaking on the mash_browsertests.
Disabling that until it can be fixed

TBR=jamescook@chromium.org
TEST=mash_browsertests

Bug:  787121 ,  678687 
Change-Id: I7805e10ac7486f0500abfdb00a092e6dc029fd46
Reviewed-on: https://chromium-review.googlesource.com/780282
Commit-Queue: Jonathan Ross <jonross@chromium.org>
Reviewed-by: Jonathan Ross <jonross@chromium.org>
Cr-Commit-Position: refs/heads/master@{#517981}
[modify] https://crrev.com/49e43eff449b72d330b32152accbb023d4622879/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Project Member

Comment 72 by bugdroid1@chromium.org, Dec 12 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c3315f3631e7004f47c7e496371ffb0c04f7b795

commit c3315f3631e7004f47c7e496371ffb0c04f7b795
Author: Steven Bennetts <stevenjb@chromium.org>
Date: Tue Dec 12 18:03:28 2017

CrSettingsBrowsertest: De-flake CrOs subpage tests

* CrSettingsInternetPageTest added flushAsync() after it was disabled
  but it was never re-enabled; it should be stable now.
* CrSettingsAndroidAppsPageTest was failing for similar reasons on
  Mash and appears to be stable with a similar fix.
* CrSettingsBluetoothPageTest follows a similar pattern so a similar
  fix was applied to prevent any potential unreported flake.

Bug:  678687 , 729607 
Change-Id: I0f1fce0a81c5b6a9c7a9f390a6761a9d341cd1f7
Reviewed-on: https://chromium-review.googlesource.com/821357
Reviewed-by: James Cook <jamescook@chromium.org>
Reviewed-by: Michael Giuffrida <michaelpg@chromium.org>
Commit-Queue: Steven Bennetts <stevenjb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#523475}
[modify] https://crrev.com/c3315f3631e7004f47c7e496371ffb0c04f7b795/chrome/test/data/webui/settings/android_apps_page_test.js
[modify] https://crrev.com/c3315f3631e7004f47c7e496371ffb0c04f7b795/chrome/test/data/webui/settings/bluetooth_page_tests.js
[modify] https://crrev.com/c3315f3631e7004f47c7e496371ffb0c04f7b795/chrome/test/data/webui/settings/cr_settings_browsertest.js
[modify] https://crrev.com/c3315f3631e7004f47c7e496371ffb0c04f7b795/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Project Member

Comment 73 by bugdroid1@chromium.org, Dec 14 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/eefdc769289d2fe0de931f6626c94aa34d0292bd

commit eefdc769289d2fe0de931f6626c94aa34d0292bd
Author: Steven Bennetts <stevenjb@chromium.org>
Date: Thu Dec 14 17:54:38 2017

AppWindowApiTest.Properties: Elim flakiness

This test was assuming that calling minimize() and maximize() followed
by restore() without waiting for onMinimized or onMaximized would
trigger onRestored. This is not the case on Mash, and is also a likely
cause of flake on other platforms.

Bug:  232330 , 678687 ,794771
Change-Id: I2a6a0f1e0b527583add555ed5ab8bfdc3a073df6
Reviewed-on: https://chromium-review.googlesource.com/821310
Reviewed-by: Ben Wells <benwells@chromium.org>
Commit-Queue: Steven Bennetts <stevenjb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#524110}
[modify] https://crrev.com/eefdc769289d2fe0de931f6626c94aa34d0292bd/chrome/test/data/extensions/platform_apps/windows_api_properties/main.js
[modify] https://crrev.com/eefdc769289d2fe0de931f6626c94aa34d0292bd/extensions/browser/api/app_window/app_window_apitest.cc
[modify] https://crrev.com/eefdc769289d2fe0de931f6626c94aa34d0292bd/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Project Member

Comment 74 by bugdroid1@chromium.org, Jan 11 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e8adc62ad8f38a703411777d851d0f4fd3924d2e

commit e8adc62ad8f38a703411777d851d0f4fd3924d2e
Author: James Cook <jamescook@chromium.org>
Date: Thu Jan 11 17:44:17 2018

cros: Enable more shelf browser tests under mash

In several cases, wait for the window manager to stabilize. For the
remaining failures, document the causes in the filter file.

Bug:  678687 
Test: browser_tests --mash
Change-Id: Id7460efa04e8f1a42c9531505a141d2148ba4c7b
Reviewed-on: https://chromium-review.googlesource.com/834886
Reviewed-by: Michael Wasserman <msw@chromium.org>
Commit-Queue: James Cook <jamescook@chromium.org>
Cr-Commit-Position: refs/heads/master@{#528663}
[modify] https://crrev.com/e8adc62ad8f38a703411777d851d0f4fd3924d2e/chrome/browser/ui/ash/launcher/chrome_launcher_controller_browsertest.cc
[modify] https://crrev.com/e8adc62ad8f38a703411777d851d0f4fd3924d2e/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Project Member

Comment 75 by bugdroid1@chromium.org, Jan 29 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9f04a28eca9fefa081c9be4872bcae249a591dd1

commit 9f04a28eca9fefa081c9be4872bcae249a591dd1
Author: James Cook <jamescook@chromium.org>
Date: Mon Jan 29 18:30:34 2018

cros: Allow ash to launch accessibility_autoclick under mash

The "accessibility_autoclick" service is a small mojo app that
implements the "automatically click when the mouse stops moving"
feature for users with difficulty clicking the mouse. It was
missing from ash's manifest file.

This fixes browser_tests AccessibilityFeatureaApiTest and
ChromeOSInfoPrivateTest.TestGetAndSet under mash.

Bug:  678687 ,  805713 
Test: browser_tests --mash
Change-Id: I3611b80209809247a64987d339cc513e1caeee06
Reviewed-on: https://chromium-review.googlesource.com/889524
Commit-Queue: James Cook <jamescook@chromium.org>
Reviewed-by: Tom Sepez <tsepez@chromium.org>
Cr-Commit-Position: refs/heads/master@{#532504}
[modify] https://crrev.com/9f04a28eca9fefa081c9be4872bcae249a591dd1/ash/manifest.json
[modify] https://crrev.com/9f04a28eca9fefa081c9be4872bcae249a591dd1/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Project Member

Comment 76 by bugdroid1@chromium.org, Jan 30 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2b7debd2fb7775b7829fa6683fc082ba774db58d

commit 2b7debd2fb7775b7829fa6683fc082ba774db58d
Author: James Cook <jamescook@chromium.org>
Date: Tue Jan 30 01:12:31 2018

cros: Disable more mash_browser_tests on mojo FYI bot

In preparation for moving the main waterfall mash_browser_tests from
its narrow whitelist to the larger FYI blacklist.

Bug:  678687 
Test: browser_tests
Change-Id: I61d512ccb1ba4152cb2ae99a022b4e8cf92c67b6
Reviewed-on: https://chromium-review.googlesource.com/891997
Reviewed-by: Michael Wasserman <msw@chromium.org>
Cr-Commit-Position: refs/heads/master@{#532713}
[modify] https://crrev.com/2b7debd2fb7775b7829fa6683fc082ba774db58d/testing/buildbot/filters/mojo.fyi.mash.browser_tests.filter

Blockedon: 807781
Blockedon: 807794
Blockedon: 810935
Project Member

Comment 80 by bugdroid1@chromium.org, Feb 14 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/cc4f20d43384571b8b57d25de4c8f9c635826d77

commit cc4f20d43384571b8b57d25de4c8f9c635826d77
Author: James Cook <jamescook@chromium.org>
Date: Wed Feb 14 19:35:33 2018

cros: Add more tests to mash_browser_tests on main waterfall

Change the filter file from a whitelist to a blacklist. The blacklist
has been in use on the Mojo FYI bots for some time and the test suite
seems stable. This adds ~8000 tests to the bot.

This is part of go/mustash. We run browser_tests --mash on the
linux-chromeos bots to test out-of-process ash sysui.

Bug:  678687 
Test: mash_browser_tests
Change-Id: I784990cb1ee4d6241961ca9744248fdbc912a231
Reviewed-on: https://chromium-review.googlesource.com/895906
Commit-Queue: James Cook <jamescook@chromium.org>
Reviewed-by: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#536775}
[modify] https://crrev.com/cc4f20d43384571b8b57d25de4c8f9c635826d77/testing/buildbot/filters/mash.browser_tests.filter

Status: Fixed (was: Started)
mash_browser_tests is on the main waterfall with 8000+ tests. Calling this done for now, since we have individual bugs for the failing tests.

We may have to revisit when service manager moves back out of the browser process, but we should see test failures then.

Components: -Internals>MUS Internals>Services>WindowService
Project Member

Comment 83 by bugdroid1@chromium.org, Apr 18 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/1a1e1a6cf983a0f5ce79d95642fbe4e41cc75bb7

commit 1a1e1a6cf983a0f5ce79d95642fbe4e41cc75bb7
Author: Evan Stade <estade@chromium.org>
Date: Wed Apr 18 20:55:09 2018

Enable SaveCardBubbleControllerImplTest on mash.browser_tests

Bug:  678687 
Change-Id: If1bc9f87184627130879775865b5e8696faf82c2
Reviewed-on: https://chromium-review.googlesource.com/1008278
Reviewed-by: James Cook <jamescook@chromium.org>
Commit-Queue: Evan Stade <estade@chromium.org>
Cr-Commit-Position: refs/heads/master@{#551818}
[modify] https://crrev.com/1a1e1a6cf983a0f5ce79d95642fbe4e41cc75bb7/testing/buildbot/filters/mash.browser_tests.filter

Project Member

Comment 84 by bugdroid1@chromium.org, Aug 22

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/0f225b910d0b7d5a590a15832c2eb8413077cb59

commit 0f225b910d0b7d5a590a15832c2eb8413077cb59
Author: Scott Violet <sky@chromium.org>
Date: Wed Aug 22 23:41:29 2018

chromeos: enables more tests for Mash

As part of https://chromium-review.googlesource.com/c/chromium/src/+/1183892
I wired up what was OutputProtection. That means these tests now pass.

BUG= 678687 
TEST=test only change

Change-Id: Ie68291b08973953e5d3d238ce87bb6d552a1b50e
Reviewed-on: https://chromium-review.googlesource.com/1185908
Reviewed-by: Xiyuan Xia <xiyuan@chromium.org>
Commit-Queue: Scott Violet <sky@chromium.org>
Cr-Commit-Position: refs/heads/master@{#585305}
[modify] https://crrev.com/0f225b910d0b7d5a590a15832c2eb8413077cb59/testing/buildbot/filters/mash.browser_tests.filter

Sign in to add a comment